Appropriate configuration when applying pull correction

[sathanas31]

"Fix" to #363:

In case of power_law epdf, if you choose static floors and pull correctors (eg median_1d) you need to provide the gamma in order to weight the MC, so basically these are applicable for fixed-gamma power law. Hence, raise an error if gamma is not provided in the llh_energy_pdf for these cases.

Other changes:

• in minimizer, raise error if try applying pull correction when choosing KDEs (ie llh_name either standard_kde_enabled or std_matrix_kde_enabled) since KDEs already account for this
• in ts_distributions, compatibility of ts_type param between plot_background_ts_distribution and fit_background_ts methods
• in ResultsHandler find_disc_potential, correct plot names when discovery potential threshold from chi2 fit and TS=25

[JannisNe]

maybe change the name to something more descriptive, like appropriate warnings for KDEs with pull-corrections